27.12.2013 Views

EA-10/16 - National Physical Laboratory

EA-10/16 - National Physical Laboratory

EA-10/16 - National Physical Laboratory

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>EA</strong>-<strong>10</strong>/<strong>16</strong> • <strong>EA</strong> Guidelines on the Estimation of Uncertainty in Hardness Measurements<br />

c<br />

i<br />

∆H<br />

≈<br />

∆x<br />

i<br />

X 1 = x1 ,..., X n = x n<br />

(2)<br />

The experimental evaluation of the sensitivity coefficients is usually<br />

time consuming, therefore usually it is advantageous to use the<br />

experimental results given in literature [4, 5] and adopted for the<br />

examples attached, but one shall be careful when the relevant<br />

factors depend on the characteristics of the material tested (dwell<br />

time and indentation velocity). In this case some experiments with<br />

the specific material are necessary.<br />

u i (H) is the contribution to the standard uncertainty associated with the<br />

hardness H resulting from the standard uncertainty u(x i ) associated<br />

with the input estimate x i :<br />

u<br />

u i (H) = c i u(x i ) (3)<br />

d) For uncorrelated input quantities the square of the standard<br />

uncertainty u(H) associated with the measured hardness H is given<br />

by:<br />

2<br />

n<br />

∑<br />

2<br />

( H ) = u i<br />

( H )<br />

(4)<br />

i=<br />

1<br />

e) Calculate for each input quantity X i the contribution u i (H) to the<br />

uncertainty associated with the hardness H resulting from the input<br />

estimate x i according to Eqs. (2) and (3) and sum their squares as<br />

described in Eq. (4) to obtain the square of the standard uncertainty<br />

u(H) of the hardness H.<br />

f) Calculate the expanded uncertainty U by multiplying the standard<br />

uncertainty u(H) associated with the hardness H by a coverage<br />

factor k=2:<br />

U = ku(H) (5)<br />

Should the effective degrees of freedom ν eff in exceptional cases be<br />

less than 15, then calculate the coverage factor k according to<br />

<strong>EA</strong>/4-02, Annex E [1].<br />

g) Report the result of the measurement as follows: in calibration<br />

certificates, the complete result of the measurement comprising the<br />

estimate H of the measurand and the associated expanded<br />

uncertainty U shall be given in the form (H±U). To this an<br />

explanatory note must be added which in the general case should<br />

have the following content:<br />

The reported expanded uncertainty of measurement has been<br />

obtained by multiplying the combined standard uncertainty by the<br />

coverage factor k=2 that, for a normal distribution, corresponds to a<br />

confidence level p of approximately 95%. The combined standard<br />

uncertainty of measurement has been determined in accordance<br />

with <strong>EA</strong>/4-02 [1].<br />

July 05 rev.00 Page 12 of 24

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!